In our final part on the area of node representation, I would like to show you how all of these

ideas, the ideas of semantic networks and ALC and so on, how they can be applied to

the semantic web, which we've also seen the intuitions of and now how this is actually

engineered.

So there's a couple of technologies we need to use and that have become developed for

actually taking the whole web as an A-box that need to be kind of adapted.

And the idea is that we do something like the assertions.

And we have two kinds of assertions.

The first assertion is A is related to B, and the other one is A is a member of a certain

concept.

So RDF has been developed to do those things at web scale.

Think of using all the object of the World Wide Web and describing them.

So shopping items, price and available if you were to do a web application like Idealo.

Time schedules, we've already talked about.

Information about web pages.

Content and writing of web pictures, search engines, electronic libraries, all of those

kind of things.

You can imagine that things easily get big.

And so you have to have a couple of ideas.

The idea is that the objects should be web resources.

And the basic idea of the web is that everything on the web, every resource has a standardized

identifier.

And that identifier is actually a URI.

So any resource has a URI.

So instance, our university has the URI typically of its home page.

So we can define something we call properties that is resources that have a name and a property

value.

And that property value might be something for name, for me, might be something like

Michael Kohl has my name.

Or it might be a URL.

So something like this URL, which happens to be my home page, which describes somehow

associated with me.

So we are, the important thing, the concept here is that of an RDF statement, which is

also known as a triple, consists of resource as defined above, namely anything that has

a URI.

We call it the subject.

A property, this thing here, like the author or so.

And a value.

So we have statements, which are triples of subject, predicate and object.

And a set of triples is called an RDF graph.

So we have a statement, for instance, this slide, which is something that is a web resource,

it does have a URI, namely it's part of quark.info slash teaching slash AI slash what is it?

Notes dot PDF and then, or slides dot PDF slash 503 has been authored, the author property

by Michael Colaza, which is either a string or URI.

The URI is the better option, of course.

So that's what RDF does.

It gives you triples.

And you can also think of triples as edges in a semantic network, for instance.

The property is the relation name and the two objects are actually the two nodes, the

edge.